Solving the problem of classification using fuzzy logic

The work is devoted to the use of fuzzy logic methods to solve problems of classification of biological objects. The construction and use of fuzzy inference by the Zadeh - Mamdani method is considered. The task is to develop logical rules for determining the sex of a viper in life according to a series of qualitative and quantitative morphological features. Two variants of solving the problem: using formal binary logic methods and using methods of fuzzy logic are shown. The main components of formal logic (concepts, judgments, laws) and fuzzy logic (membership functions, linguistic variables, calculation of fuzzy inferences) are considered. The methods of defining membership functions are indicated; the method of composing logistic equations using distributions of feature values over an extensive sample of animals is implemented. It is proposed to interpret membership functions as a semantic filter that enhances the diagnostic properties of the studied features. All calculations are illustrated by numerical examples. For all stages of modeling, their own scripts are written, attached by hyperlink to the text. Executing scripts in the R package environment will allow getting acquainted with all the stages of the study in detail. The reasons that this method is not in demand in the practice of environmental research are discussed. The issues of parameter settings and verification methods of the logical model are considered. It is noted that the direction of integration of fuzzy logic with neural network modeling is actively developing.